Membrane-Based, Liquid–Liquid Separatorwith Integrated Pressure Control

The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters.

Citation Adamo, Andrea, Patrick L. Heider, Nopphon Weeranoppanant,and Klavs F. Jensen. “Membrane-Based, Liquid–Liquid Separatorwith Integrated Pressure Control.” Ind. Eng. Chem. Res. 52, no. 31(August 7, 2013): 10802–10808.

As Published http://dx.doi.org/10.1021/ie401180t

Publisher American Chemical Society (ACS)

Version Author's final manuscript

Citable link http://hdl.handle.net/1721.1/92770

Terms of Use Creative Commons Attribution-Noncommercial-Share Alike

Detailed Terms http://creativecommons.org/licenses/by-nc-sa/4.0/

1

Membrane-Based, Liquid-Liquid Separator with

Integrated Pressure Control

Andrea Adamo,† Patrick L. Heider,† Nopphon Weeranoppanant, and Klavs F. Jensen*

Department of Chemical Engineering

Massachusetts Institute of Technology

77 Massachusetts Avenue

Cambridge, MA 02139, USA

2

ABSTRACT

We describe the development and application of an improved, membrane-based, liquid-liquid

separator. Membrane based separation relies on the exploitation of surface forces and the use of a

membrane wetted by one of the phases; however, successful separation requires accurate control

of pressures making the operation and implementation cumbersome. Here we present an

improved separator design that integrates a pressure control element to ensure that adequate

operating conditions are always maintained. Additionally, the integrated pressure control

decouples the separator from downstream unit operations. A detailed examination of the

controlling physical equations shows how to design the device to allow operation across a wide

range of conditions. Easy to implement, multistage separations such as solvent swaps and

countercurrent extractions are demonstrated. The presented design significantly simplifies

applications ranging from multistep synthesis to complex multistage separations.

INTRODUCTION

The last decade has seen a growing effort in the development of continuous flow chemical

systems ranging from microscale1-7 to milliscale.8-13 The field has expanded in scope beyond

single reactor and synthesis steps to multistep synthesis including intermediate workup and

separation steps.14-19 Among the separation techniques available to organic chemists, liquid-

liquid separations are becoming more popular.20 The small length scales encountered in

continuous systems on the microscale and milliscale improve the extraction rate due to higher

mass transfer coefficients,21, 22 but present challenges for separation as surface forces dominate

over the traditionally used gravity force.23 Several solutions have reported including parallel

3

flow,24-26 settling tanks,27, 28 selectively wetting channels,29-31 centrifugation,13, 32 and

microfiltration membranes.19, 33 All of these separation techniques (except settling tanks and

centrifugation) rely on the interfacial tension, γ, between the two liquids to provide the force to

achieve separation.

Previous studies have demonstrated capillary pressure, Pcap, is a critical parameter which must

be balanced with other forces in the device to guarantee proper operation of a separator.19, 29, 33, 34

Pcap is given as

2 cos

capPr

(1)

where θ is the contact angle between the solid material of the device and two liquid phases and r

is the radius of curvature of the interface. The design of each system is such that Pcap must be

higher than the pressure difference across the interface, or separation will be incomplete. Since γ

and θ are material properties that vary for each application, the only available design parameter

is r which must be minimized to produce stable operation. In microscale systems, this is easy to

do in a single channel with a small gap for parallel flow or two selectively wetting channels.

Milliscale systems require larger channels to keep the pressure drop small relative to Pcap and

therefore need many smaller channels to maintain high Pcap while permitting high flow rates.

This is easily achieved using microfiltration membranes.19, 33 Membrane-based separators are

therefore ideal candidates for operating at both microscale16 and milliscale.19

Previous analyses of membrane-based separators have focused on the importance of balancing

the flow resistances through each outlet33 or pressure control of individual outlets.19, 29 These

design criteria produce functioning devices; however, they make integration into larger systems

difficult as downstream fluctuations can easily disrupt separation. A solution to this problem has

been the use of break tanks and additional pumps16 which increases the volume and complexity

4

of the systems. Break tanks with additional pumps can be eliminated by careful design for

specific operating conditions at the cost of flexibility and robustness. Direct control of the

pressure at the separator outlets can also avoid break tanks,19 but still requires a feedback loop to

account for differences in operating conditions downstream.

This paper reexamines the governing equations for separation in a membrane-based separator

and presents a new separator design with an integrated pressure control component. The new

design decouples the unit from downstream operations, increases flexibility, and eliminates the

necessity for feedback control. Easy to implement solvent swap and countercurrent extraction

examples are presented.

DESIGN THEORY

There are two main failure modes for membrane-based separators (Figure 1). The first is

breakthrough of the retained phase which occurs when the pressure difference across the

membrane (transmembrane pressure, ΔPmem) is greater than Pcap (Figure 1b). The second failure

mode is when the permeate phase is partially retained by the membrane and exits with the

retained phase (Figure 1c). This occurs when there is insufficient pressure to cause the permeate

liquid to flow through the membrane (Pper) which can be approximated by

4

8per

QLP

n R

(2)

where μ is the viscosity of the permeate phase, Q is the entering permeate liquid volumetric flow

rate, L is the membrane thickness, n is the number of pores and R is the pore radius. This

assumes that the membrane acts as an array of cylindrical pores which is acceptable for this

analysis as shown below. The value of Pper is the minimum pressure required to drive all of the

permeate phase through the membrane using the entire area. Separation will still occur at values

5

of ΔPmem greater than Pper, but the entire membrane area is not utilized (n is smaller such that

Equation (2) equals ΔPmem). In this case, the non-active pores support the additional pressure

using the interfacial tension force. This is possible for values of ΔPmem up to the point when

breakthrough occurs, described as the first failure mode. A third failure mode exists where two

phase streams exit both outlets, but this is indicative of operating the separator at a flow rate

excessive for the available membrane area and therefore this condition is generally not

encountered.

When the pressure drop along the length of the membrane channel is negligible compared to

Pcap – Pper, then these two failure modes can be described in a single compound inequality

cap mem perP P P (3)

since ΔPmem can be assumed to be constant along the membrane. This assumption is satisfied

when the channel is sized so that pressure drops along its length are negligible. The first

inequality is satisfied by appropriate selection of the membrane (material and pore size) while

the second is better understood by replacing ΔPmem with P1 – P2 where P1 is the pressure on the

retentate side of the membrane and P2 is the pressure on the permeate side of the membrane and

rewriting as

2 1 2cap perP P P P P (4)

Equation (4) means that successful designs must always operate under conditions where the

retentate pressure is some value greater than the sum of the permeate pressure and the Pper value

for the maximum flow rate desired through the membrane.

Figure 1 shows a schematic of the membrane separator which incorporates a pressure control

segment immediately following the membrane. The pressure control is made up of a diaphragm

stretched over the retentate stream with the permeate stream flowing on the reverse side. Since

6

the diaphragm seals against the retentate flow path, no flow exits the retentate side of the

separator unless P1 > P2. Additionally, the membrane is slightly deformed conveying an

additional force on the retentate flow path that must be exceeded to permit flow. The differential

pressure applied increases with the amount of deformation, the thickness of the diaphragm, and

the elastic modulus of the diaphragm. This device then acts as a differential pressure controller

such that P1 = Pdia + P2 where Pdia is the additional pressure due to the tension on the diaphragm.

This simplifies Equation (4) to

cap dia perP P P (5)

which means that an appropriately designed separator will achieve complete separation as long

as the flow rate through the membrane remains below maximum value for the design (that is

pressure drops along channels are negligible).

EXPERIMENTAL SECTION

Figure 2 shows a photograph of a separator constructed out of polycarbonate. Better chemical

compatibility was provided by constructing the separator out of high molecular weight

polyethylene (HDPE). HDPE was selected because it was compatible with the solvents tested

and did not deform or degrade. Even higher chemical compatibility could be provided by using

materials such as ethylene tetrafluoroethylene (ETFE). The parts in the device were constructed

using conventional machining tools and assembled with off the shelf screws and O-rings. The

membranes used were Pall Zefluor 1 μm PTFE microfiltration membranes which are wet by the

permeating organic phase and retain the aqueous phase with a total area of 280 mm2 (roughly 35

mm long by 8 mm wide by 1 mm high). A PTFE membrane was selected because of its high

chemical compatibility. No change in membrane performance was observed with different

7

a

b

c

Figure 1. Integrated pressure control in membrane separator showing deformed diaphragm

(heavy curved line) to provide a fixed pressure difference across the membrane (short vertical

lines). The aqueous and organic phases are shown in blue and white respectively. (a) Separator

under normal operation. (b) Separator operating with breakthrough of the retained phase. (c)

Separator operating with retention of the permeate phase.

solvents due to mechanical changes in the membrane. The diaphragm was made of

perfluoroalkoxy (PFA) film with a thickness of 25 μm and 50 μm. The 25 μm film provided a

smaller pressure difference, but was prone to damage so the thicker film was used which was

8

more robust but provided a higher pressure difference (Figure 7). PFA was the only material

tested for the diaphragm because it provides high chemical compatibility across a range of

compounds commonly encountered in organic synthesis while other materials (such as various

rubbers and elastomers) are susceptible to degradation in the presence of certain chemicals. All

pumping was performed using piston pumps from Knauer (Smartline pump 100), Eldex (Optos

2SIP), and Fuji Techno Industries (Super metering pump HYM-08).

Outlets

Inlet

10 mm

Figure 2. Photograph of polycarbonate membrane separator with integrated pressure control.

The separator membrane is located on the lower portion and the pressure control diaphragm is

located on the upper portion of the device.

The design equations were tested by removing the loops connecting the separation membrane

with the differential pressure controller and attaching 235 cm of 1.6 mm inner diameter (ID)

tubing to the permeate outlet and 60 cm of 0.76 mm ID tubing and 183 cm of 1.6 mm ID tubing

to the retentate outlet. Previously contacted and separated water and ethyl acetate were each

pumped at 5 mL/min into a tee mixer and then passed into the membrane separator. Pressure

variations across the membrane were controlled by varying the height of the outlets over 2 m and

calculating the pressure at the membrane. Failure points were calculated using Equation (3). The

9

values for R and L were taken from the manufacturer’s specifications and n was calculated by

measuring the membrane flow resistance by flowing only toluene and measuring the flow rate

split between the two outlets. The value of γ was taken from the literature35 as 36.1 mN/m for

toluene-water and 6.8 mN/m for ethyl acetate-water. The differential pressure controller was

tested by measuring the flow rate split between the two outlets when flowing only toluene

through the device. 50 cm of 0.51 mm ID tubing was added to increase the pressure drop on the

permeate side of the membrane so that flow exits both outlets.

Characterization of the new separator design where the membrane is coupled to the pressure

controller was done following the set up represented in Figure 3. Hexane-water and ethyl

acetate-water pairs were tested. The first one provided an initial test bed for the system with a

wide operating range due to a high interfacial tension (50 mN/m), the second was a more

challenging separation as the interfacial tension between the two fluids is an order of magnitude

lower (6.8 mN/m) thus restricting the pressure difference operating window for successful

separation. Flow rates between 2 and 8 mL/min for both the aqueous and organic phase were

tested with these solvents as shown in Figure 3. Additionally to test the robustness of the

pressure controller, the pressure in the collection reservoir of the organic side was changed to

apply backpressure to the separator simulating downstream pressure drop (0, 1, 1.4, and 2 bar

were used).

10

Figure 3. Setup for testing of membrane separator with integrated pressure control. Organic

phase lines are shown in dark red, water lines are shown in blue, and two phase streams are

shown as dashed lines. The organic outlet of the separator entered a closed vessel pressurized by

a gas.

Solvent swap

A two stage solvent swap was tested as shown in Figure 4. A stream of 0.34 M benzoic acid in

ethyl acetate was pumped at 1 mL/min into a mixing tee to contact a stream of 0.55 M NaOH

flowing at 1 mL/min. A HDPE separator then split the phases with the organic phase passing

through the membrane and through an additional 0.6 bar backpressure controller. This was added

because the organic outlet of the separator generally must be at a higher pressure than the

aqueous or else excess pressure on the aqueous outlet (in this case due to the second separator)

will cause breakthrough of the aqueous phase by overriding the differential pressure controller.

The aqueous phase containing benzoic acid then contacted a 1 mL/min stream of toluene. After a

short length of tubing, a 1 mL/min stream of 0.6 M HCl was added. The stream was then

separated by a second, identical membrane separator. Tubing with lengths to provide > 10 s of

residence time was used to allow the streams to reach equilibrium prior to separation. All tubing

used was 1.6 mm ID. The flow rate of each outlet stream was measured by collection in

graduated cylinders and the concentration of benzoic acid determined by HPLC (Agilent 1100

11

with UV detector, 30 mM H3PO4 aqueous mobile phase, and 1:1 acetonitrile:methanol organic

mobile phase). The results were compared to a comparable batch extraction performed at a 50

mL scale in separatory funnels with careful measurement of volumes.

Figure 4. Flow diagram of solvent swap setup. Organic phase lines are shown in dark red, water

lines are shown in blue, and two phase streams are shown as dashed lines.

Countercurrent extraction

A three stage countercurrent extractor was set up as shown in Figure 5. Toluene and water

were used as solvents with acetone as the extractant added at a 0.05 mass ratio to either solvent.

The toluene feed was pumped by a single pump while a separate aqueous phase pump was used

at each stage to increase the pressure so countercurrent operation was possible. All pumps were

set to 3 mL/min. All tubing was 0.76 mm ID PFA with sufficient lengths to allow equilibrium

before separation at each stage (residence time > 5 s). The outlet flow rates were measured and

the concentration of acetone was determined by HPLC (Agilent 1100 with RI detector and 5 mM

H2SO4 isocratic mobile phase) for the aqueous phase and GC (HP 6890 with FID detection) for

the organic phase. Samples were taken after 15 min of operation with 3-5 repeats over the first

hour. Performance was determined using standard countercurrent extraction plots.36

12

Figure 5. Flow diagram of countercurrent extraction setup. Organic phase lines are shown in

dark red, water lines are shown in blue, and two phase streams are shown as dashed lines.

RESULTS AND DISCUSSION

The initial design of the separator was tested to confirm the assumptions from the theory

section above held for the separator. The separator was operated without the differential pressure

controller at 5 mL/min with a 1:1 ethyl acetate-water system. The results shown in Figure 6

compare the model predictions of the flow rate out the permeate side of the membrane with

experimental results for varying ΔPmem. While the performance of the separator is qualitatively

similar to the model, breakthrough of the aqueous phase happens at much lower ΔPmem values.

This is indicative of a pore size distribution where a small number of pores with larger radii

allow the aqueous phase through while the other pores still prohibit flow. This result is similar to

13

a bubble point test. A bubble point test, however, typically only identifies a critical pressure,

converted to pore size using Equation (1), while this result shows a gradual increase in

permeation as ΔPmem is increased. This demonstrates that even if the largest pores are

compromised, complete loss of separation will not occur as the low number of large pores cannot

support the full flow of the retained phase. While Equation (5) can be expressed in terms of pore

size distribution, separation should occur as long as the Pcap for the largest pores (determined

empirically) is heeded. The large flat region in Figure 6 shows that the design parameters

described above can be used to select an appropriate Pdia for the differential pressure controller to

allow separation by operating below the empirically determined Pcap. Figure 7 shows that the

differential pressure controller performance across a wide range of flow rates is fairly constant

and within an acceptable range for most applications. The values of Pcap for both the

toluene/water and ethyl acetate/water systems and the values of Pper for each flow rate bound the

values for Pdia meaning the separator satisfies Equation (5) and should separate effectively.

Figure 7 shows that even with increasing flow rate, Pdia, which equals ΔPmem in the device, will

remain constant and therefore separation performance remains constant. This is not true when the

differential pressure controller is not present and downstream pressure drops change due to

varying flow rates.

14

00.20.40.60.8

11.21.41.61.8

2

0 0.1 0.2 0.3Nor

mal

ized

per

mea

te f

low

rat

e

ΔPmem (bar)

Figure 6. Plot of model and experimental results of the membrane separator without the pressure

control diaphragm. The flow rate through the permeate outlet normalized by the inlet organic

flow rate is plotted versus ΔPmem where a value of 1 means perfect separation. Model values are

given by the solid line while experimental values are given by circles. A flow rate of 5 mL/min

for both water and ethyl acetate was used.

0.01

0.1

1

10

0 20 40 60

Pre

ssu

re (

bar

)

Flow rate (mL/min)

Figure 7. Plot of the differential pressure controller’s performance, Pdia, versus flow rates

(circles). The solid and dashed lines represent Pcap for a toluene-water and ethyl acetate-water

separation respectively. The dotted line represents Pper for each flow rate. The separator meets

the criteria of Equation (5) for all flow rates.

15

Results from the testing of the operation of a single separator with the integrated pressure

controller are shown in Table 1. The table shows that as back pressure increases, a modest

breakthrough can be observed for the ethyl acetate-water pair. This can be interpreted in light of

the results of Figures 6 and 7. Figure 6 shows that the ethyl acetate-water system begins to allow

breakthrough of the aqueous phase at 0.15 bar while Figure 7 shows that Pdia was between 0.1

and 0.15 bar for single phase flow. High pressures applied to the organic outlet cause small

deviations in the performance of the pressure control diaphragm, slightly increasing Pdia which

was sufficient to cause some breakthrough in the ethyl acetate-water system. Importantly, the

hexane-water system has a sufficiently high Pcap that the small increase in did not cause any

failure in separation across all the conditions tested.

Table 1. Summary of performance of a membrane separator with integrated pressure control

Aqueous flow rate

Organic flow rate

Normalized permeate flow ratea Backpressure on organic outlet

mL/min mL/min 0 bar 1 bar 1.4 bar 2 bar

Hex

ane-

wat

er

2 2 1 1 1 1

5 5 1 1 1 1

8 8 1 1 1 1

2 8 1 1 1 1

8 2 1 1 1 1

Eth

yl a

ceta

te-w

ater

2 2 1 1 1.0425 1.0500

5 5 1 1.0108 1.0384 1.0930

8 8 1 1.0159 1.0476 1.1180

2 8 1 1 1.0625 1.2420

8 2 1 1.0213 1.0492 1.0123

aThe flow rate through the permeate outlet is normalized by the inlet organic flow rate (where 1 is perfect separation).

16

Solvent swap

The ability to change solvents between successive synthesis steps (‘solvent swap’) is an

important consideration in using optimum solvent choices for the individual reactions. As a

demonstration case, the separator was tested in a two stage solvent swap of benzoic acid from

ethyl acetate to toluene. The results of the extraction are summarized in Table 2. The continuous

system reproduces the batch performance both in terms of yield and mass balance. The system

was operated for 2 h (corresponding to over 60 residence volumes) without any failure in the

separation and halted without failure of the separation. No additional control was required

beyond starting each pump up sequentially (the first stage pumps, then the second stage pumps).

This application demonstrated how the separator effectively decouples the membrane pressures

from effects downstream when the pressure on the organic outlet is greater than the aqueous

outlet. The 0.6 bar of backpressure on the first stage permeate (organic) outlet would normally

prevent flow without the presence of the differential pressure controller. Without the differential

pressure control diaphragm, the backpressure would prevent all flow through the membrane and

halt any separation. This is similar to the effect when a separator is used to separate an organic

phase which contains the reactant for a subsequent reaction. Normally increased pressure

required for a second reactor (to increase the boiling point of a solvent or due to a fixed bed)

would require additional pressure control on the retentate (aqueous) side to closely match the

organic pressure. The differential pressure controller simplifies the system by negating the need

for the retentate side pressure control.

17

Table 2. Summary of solvent swap results comparing a batch shake flask and continuous with

two membrane separators

Yielda Mass balancea

Continuous 0.92 +/- 0.006 0.97 +/- 0.008

Shake flask 0.94 0.96

aError values are one standard deviation of 4 samples taken over the 2 h run.

A similar situation existed on the retentate side of the first separator in that it fed a second

separator with a differential pressure controller that elevated the pressure. The effects of this

pressure on the first separator were eliminated by adding a static backpressure regulator to the

permeate side of the first separator larger than the highest pressure the retentate side might see. If

this was not done, the first separator will not separate at all, but after adding the fixed regulator

no additional control was required as long as the retentate pressure remained below 0.6 bar. The

first stage was completely isolated from the second stage such that even during startup when the

second stage was not running, the first stage will separate both streams. No additional

adjustments were needed to maintain separation when the second stage was started up.

Countercurrent extraction

The separators were also tested in a countercurrent extraction setup. This required setting up

one pump for the organic phase and a pump for the aqueous phase for each stage to drive the

countercurrent flow. The separator achieved three stages of extraction (Figure 8) when extracting

acetone from either toluene or water. This was expected since the high mass transfer rates in slug

flow21, 22 ensured that equilibrium was achieved before separation at each stage. Since extraction

and separation occur independently, this system could be scaled to a large number of stages with

18

0

0.01

0.02

0.03

0.04

0.05

0.06

0 0.02 0.04 0.06

Mas

s ra

tio

of a

ceto

ne

in t

olu

ene

Mass ratio of acetone in water

a

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.00 0.02 0.04 0.06

Mas

s ra

tio

of a

ceto

ne

in t

olu

ene

Mass ratio of acetone in water

b

Figure 8. Extraction diagram for (a) extraction of acetone from water into toluene and (b)

extraction of acetone from toluene into water. The equilibrium curve is shown by a dotted line

and the operating line is shown by the dashed line; stages are stepped off with a solid line.

additional separators and pumps without any extraction efficiency issues. This is an

improvement over other small scale countercurrent systems24, 37, 38 which have a low number of

stages and are limited in their flow rate by stage efficiencies related to the throughput. The

19

decoupling of the pressure in each stage makes the setup and operation of the countercurrent

separator simple. The system needed to be initially primed with solvent for each phase as the

piston pumps used could not handle two phase streams. After turning on each pump, the system

then operated without any outside control for over 1 h, eliminating level control between stages

for mixer/settler systems.27

CONCLUSIONS

The membrane-based separator presented here is an advance in continuous flow separators at

the milliscale. The integrated pressure controller greatly reduces the complexity when

implementing separators within chemical reactor systems by decoupling the separator pressures

and limiting the online control required for operation. It is the first demonstration of

countercurrent multi-stage liquid-liquid extraction using membrane based separators. Further, it

simplifies multistep chemical systems including reactions coupled to separation and other

multistage separations such as cross flow extraction. Larger systems which incorporate dynamic

control would benefit from the intrinsic pressure control since control actions which affect

downstream pressures (flow rate fluctuations, composition changes, etc.) can be tolerated

without redesigning the system or additional control on the separator. The separator is useful in

research settings since little additional changes are required to switch the separator between

different applications. The specific design is especially suited to flow rates between 1 and 10

mL/min. Lower flow rates are possible although the residence time increases so a smaller device

may be more appropriate depending on the application. Higher flow rates were also achieved (up

to 40 mL/min) for systems with higher interfacial tension, limited by the pressure drop in the

membrane channel compared to Equation (2). Future work to scale the separator down to μL/min

20

flows and determine the upper limits for scale up to larger separators using this design are

ongoing as well as applications to more complex, multistep syntheses.

AUTHOR INFORMATION

Corresponding Author

* E-mail: kfjensen@mit.edu. Tel.: +1-617-253-4589. Fax: +1-617-258-8224.

Author Contributions

† These authors contributed equally.

Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENT

We would like to acknowledge the Novartis-MIT Center for Continuous Manufacturing and

DARPA Grant #N66001-11-C-4147 for their generous funding of this research.

REFERENCES

(1) Webb, D.; Jamison, T. F., Continuous flow multi-step organic synthesis. Chem. Sci. 2010, 1, (6), 675-680. (2) Jensen, K. F., Microreaction engineering -- is small better? Chem. Eng. Sci. 2001, 56, (2), 293-303. (3) Wiles, C.; Watts, P., Recent advances in micro reaction technology. Chem. Commun. 2011, 47, (23), 6512-6535. (4) Hartman, R. L.; McMullen, J. P.; Jensen, K. F., Deciding Whether To Go with the Flow: Evaluating the Merits of Flow Reactors for Synthesis. Angew. Chem. Int. Ed. 2011, 50, (33), 7502-7519. (5) Pennemann, H.; Watts, P.; Haswell, S. J.; Hessel, V.; Löwe, H., Benchmarking of Microreactor Applications. Org. Process Res. Dev. 2004, 8, (3), 422-439. (6) Kockmann, N.; Gottsponer, M.; Zimmermann, B.; Roberge, D. M., Enabling Continuous-Flow Chemistry in Microstructured Devices for Pharmaceutical and Fine-Chemical Production. Chem. Eur. J. 2008, 14, (25), 7470-7477.

21

(7) Wegner, J.; Ceylan, S.; Kirschning, A., Ten key issues in modern flow chemistry. Chem. Commun. 2011, 47, (16), 4583-4592. (8) Battilocchio, C.; Baumann, M.; Baxendale, I. R.; Biava, M.; Kitching, M. O.; Ley, S. V.; Martin, R. E.; Ohnmacht, S. A.; Tappin, N. D. C., Scale-Up of Flow-Assisted Synthesis of C2-Symmetric Chiral PyBox Ligands. Synthesis 2012, 2012, (04), 635-647. (9) McMullen, J. P.; Jensen, K. F., Rapid Determination of Reaction Kinetics with an Automated Microfluidic System. Org. Process Res. Dev. 2011, 15, (2), 398-407. (10) Cervera-Padrell, A. E.; Nielsen, J. P.; Jønch Pedersen, M.; Müller Christensen, K.; Mortensen, A. R.; Skovby, T.; Dam-Johansen, K.; Kiil, S.; Gernaey, K. V., Monitoring and Control of a Continuous Grignard Reaction for the Synthesis of an Active Pharmaceutical Ingredient Intermediate Using Inline NIR spectroscopy. Org. Process Res. Dev. 2012, 16, (5), 901-914. (11) Kockmann, N.; Gottsponer, M.; Roberge, D. M., Scale-up concept of single-channel microreactors from process development to industrial production. Chem. Eng. J. 2011, 167, (2–3), 718-726. (12) Hessel, V.; Löb, P.; Löwe, H., Industrial Microreactor Process Development up to Production. In Microreactors in Organic Synthesis and Catalysis, Wiley-VCH Verlag GmbH & Co. KGaA: 2008; pp 211-275. (13) Schuur, B.; Hallett, A. J.; Winkelman, J. G. M.; de Vries, J. G.; Heeres, H. J., Scalable Enantioseparation of Amino Acid Derivatives Using Continuous Liquid−Liquid Extraction in a Cascade of Centrifugal Contactor Separators. Org. Process Res. Dev. 2009, 13, (5), 911-914. (14) Smith, C. J.; Nikbin, N.; Ley, S. V.; Lange, H.; Baxendale, I. R., A fully automated, multistep flow synthesis of 5-amino-4-cyano-1,2,3-triazoles. Org. Biomol. Chem. 2011, 9, (6), 1938-1947. (15) Bogdan, A. R.; Poe, S. L.; Kubis, D. C.; Broadwater, S. J.; McQuade, D. T., The Continuous-Flow Synthesis of Ibuprofen. Angew. Chem. Int. Ed. 2009, 48, (45), 8547-8550. (16) Sahoo, H. R.; Kralj, J. G.; Jensen, K. F., Multistep Continuous-Flow Microchemical Synthesis Involving Multiple Reactions and Separations. Angew. Chem. Int. Ed. 2007, 46, (30), 5704-5708. (17) Varas, A. C.; Noël, T.; Wang, Q.; Hessel, V., Copper(I)-Catalyzed Azide–Alkyne Cycloadditions in Microflow: Catalyst Activity, High-T Operation, and an Integrated Continuous Copper Scavenging Unit. ChemSusChem 2012, 5, (9), 1703-1707. (18) Hartman, R. L.; Naber, J. R.; Buchwald, S. L.; Jensen, K. F., Multistep Microchemical Synthesis Enabled by Microfluidic Distillation. Angew. Chem. Int. Ed. 2010, 49, (5), 899-903. (19) Cervera-Padrell, A. E.; Morthensen, S. T.; Lewandowski, D. J.; Skovby, T.; Kiil, S.; Gernaey, K. V., Continuous Hydrolysis and Liquid–Liquid Phase Separation of an Active Pharmaceutical Ingredient Intermediate Using a Miniscale Hydrophobic Membrane Separator. Org. Process Res. Dev. 2012, 16, (5), 888-900. (20) Tzschucke, C. C.; Markert, C.; Bannwarth, W.; Roller, S.; Hebel, A.; Haag, R., Modern Separation Techniques for the Efficient Workup in Organic Synthesis. Angew. Chem. Int. Ed. 2002, 41, (21), 3964-4000. (21) Kuhn, S.; Jensen, K. F., A pH-Sensitive Laser-Induced Fluorescence Technique To Monitor Mass Transfer in Multiphase Flows in Microfluidic Devices. Ind. Eng. Chem. Res. 2012, 51, (26), 8999-9006.

22

(22) Burns, J. R.; Ramshaw, C., The intensification of rapid reactions in multiphase systems using slug flow in capillaries. Lab Chip 2001, 1, (1), 10-15. (23) Günther, A.; Jensen, K. F., Multiphase microfluidics: from flow characteristics to chemical and materials synthesis. Lab Chip 2006, 6, (12), 1487-1503. (24) Aota, A.; Nonaka, M.; Hibara, A.; Kitamori, T., Countercurrent Laminar Microflow for Highly Efficient Solvent Extraction. Angew. Chem. Int. Ed. 2007, 46, (6), 878-880. (25) Aota, A.; Mawatari, K.; Kitamori, T., Parallel multiphase microflows: fundamental physics, stabilization methods and applications. Lab Chip 2009, 9, (17), 2470-2476. (26) Aota, A.; Mawatari, K.; Takahashi, S.; Matsumoto, T.; Kanda, K.; Anraku, R.; Hibara, A.; Tokeshi, M.; Kitamori, T., Phase separation of gas–liquid and liquid–liquid microflows in microchips. Microchim. Acta 2009, 164, (3), 249-255. (27) Hu, D. X.; O’Brien, M.; Ley, S. V., Continuous Multiple Liquid–Liquid Separation: Diazotization of Amino Acids in Flow. Org. Lett. 2012, 14, (16), 4246-4249. (28) O'Brien, M.; Koos, P.; Browne, D. L.; Ley, S. V., A prototype continuous-flow liquid-liquid extraction system using open-source technology. Org. Biomol. Chem. 2012, 10, (35), 7031-7036. (29) Castell, O. K.; Allender, C. J.; Barrow, D. A., Liquid-liquid phase separation: characterisation of a novel device capable of separating particle carrying multiphase flows. Lab Chip 2009, 9, (3), 388-396. (30) Kashid, M. N.; Harshe, Y. M.; Agar, D. W., Liquid−Liquid Slug Flow in a Capillary: An Alternative to Suspended Drop or Film Contactors. Ind. Eng. Chem. Res. 2007, 46, (25), 8420-8430. (31) Peroni, D.; van Egmond, W.; Kok, W. T.; Janssen, H.-G., Advancing liquid/liquid extraction through a novel microfluidic device: Theory, instrumentation and applications in gas chromatography. J. Chromatogr. A 2012, 1226, (0), 77-86. (32) Schuur, B.; Floure, J.; Hallett, A. J.; Winkelman, J. G. M.; deVries, J. G.; Heeres, H. J., Continuous Chiral Separation of Amino Acid Derivatives by Enantioselective Liquid−Liquid Extraction in Centrifugal Contactor Separators. Org. Process Res. Dev. 2008, 12, (5), 950-955. (33) Kralj, J. G.; Sahoo, H. R.; Jensen, K. F., Integrated continuous microfluidic liquid-liquid extraction. Lab Chip 2007, 7, (2), 256-263. (34) Hartman, R. L.; Sahoo, H. R.; Yen, B. C.; Jensen, K. F., Distillation in microchemical systems using capillary forces and segmented flow. Lab Chip 2009, 9, (13), 1843-1849. (35) Donahue, D. J.; Bartell, F. E., The Boundary Tension at Water-Organic Liquid Interfaces. J. Phys. Chem. 1952, 56, (4), 480-484. (36) Seader, J. D.; Henley, E. J., Separation Process Principles. 2nd ed.; John Wiley & Sons, Inc.: 2006. (37) Lam, K. F.; Cao, E.; Sorensen, E.; Gavriilidis, A., Development of multistage distillation in a microfluidic chip. Lab Chip 2011, 11, (7), 1311-1317. (38) Zhang, Y.; Kato, S.; Anazawa, T., Vacuum membrane distillation by microchip with temperature gradient. Lab Chip 2010, 10, (7), 899-908.

23

TABLE OF CONTENTS GRAPHIC